The 6 Questions You Should Be Asking Your Marketing Analytics Vendor

The 6 Questions You Should Be Asking Your Marketing Analytics Vendor

truesight logoAnyone who tells you that Marketing Analytics is an exact science is wrong. The algorithms, the data, the optimizations, and the simulations can all lead to many different outcomes depending on who and how it is processed and measured. When done correctly, there is much to gain. One can optimize budgets and increase Returns On Investment (ROI), predict Sales across Product Categories, and grow your business significantly. On the flip side, however, when done incorrectly, there is much to be lost – like misdirected investment shifts and decline in channel performance.

As such, choosing an Analytics partner that is the right fit for your business is an important task. The range of companies, services, and software being offered are quite diverse; they will collect different data, crunch the data through different software, and find different answers. Not all analytic solutions will be the right fit for your business so it’s important to vet your options thoroughly.

Read more: 5 Use Cases for Natural Language Processing (NLP) Techniques in Marketing Analytics

Aside from the typical questions – what services are offered, what analytical approach is taken, and the ability to scale – central to your vendor search, we highlight 6 not-so-obvious questions to ask your analytics vendor in order to help you make a more informed opinion.

1. How Transparent Are You in Sharing Your Modelling Approach?

Transparency is not to be overlooked when it comes to Marketing Analytics. Understanding how your solutions provider built the model and how it works will allow you to better understand its nuances, limitations, and outcomes. Your providers’ willingness to explain in detail their approach and “open the box” behind the model will show a lot about their company culture and approach they will take in working with you. Consider researching whether the vendor has published their methodology work in a peer-reviewed academic journal if you are doubtful of their answer.

2. If This Goes Wrong, Will You Be Able to Pinpoint Why?

Although it’s common to be reassured that the likelihood of things going wrong is slim, it’s important nonetheless to know what the provider’s game-plan would be for mitigating such risk. The simple truth is that a lot of Analytics program can and will fail, and there needs to be bumpers in place to mitigate this. Providers should focus on the solution to your company’s problem, not on some magic tool that promises to do everything. They should also be diligent with the data – taking both collection and validation seriously. Finally, providers should guide the team in getting cross-functional buy-in from all key stakeholders consistently throughout the entire process.

3. Do You Measure Long-Term Brand Effects in Your Models?

Long-term effects in Advertising and Marketing refer to the power that Advertising holds to influence consumer valuation and behavior over the long-term. This scenario increases a consumer’s likelihood of purchase as well as how much they will pay for said product. Long-term effects aren’t traditionally accounted for in MMM models as they don’t measure the persistent time-series impact of Advertising over time.

Less savvy marketers also aren’t as concerned with these effects as they don’t have an immediate impact on brand and sales performance. As a brand marketer, you must decide whether accurately measuring long-term effects are important to your brand – we posit that they hold quite a lot of power – and how much of an impact they might have on your ROI both now and in the future.

4. Are You a Private or VC Funded Company?

When a company is funded with VC money, the number of competing stakeholders and P&L’s quadruples. With so many players on the field, the overarching goal and company approach can get lost. It’s also often the promise of scalable, automated, software that attracts VC investors to the company in the first place. The issue with this one-size-fits-all software trend, however, is that attribution is not a one-size-fits-all solution! When companies are pushed to develop these scalable solutions, their bespoke advice and industry-unique approach gets lost, which can leave you largely without actionable results. A private company unencumbered by competing priorities is a reassurance that your goals come first.

5. Do Most of Your Customers Use You For MTA, MMM, Or Both Equally?

Most Analytics companies will offer and talk about both services which is why this question might not be a priority. However, the analytics agency preferred service from clients is an important differentiator. An agency that specializes primarily in MMM, with less than a handful of clients using MTA, will be focused on different goals with a team that is structured differently from an agency more focused on MTA.

6. How Many Case Studies Do You Have?

A fancy demo might do a good job of selling software, but once that software is in your hands, how can you use it to drive real, actionable results for your business? Case studies are a great way to understand how other businesses used the vendor and platform effectively, plus they give you an idea of how deep you can dive with your insights. If a company has a good-looking demo and no case studies to boot, be weary of what this software can do for your organization.

There’s a lot that goes into a detailed, well-executed analytics model, and vetting your provider is an important part of this process. A crucial step before one begins this screening would be to sit down with key stakeholders and determine the priorities from a cultural and working perspective. Once you know where you collectively stand, you can begin the fun part – shopping around for the perfect vendor and asking the questions that matter.

Read more: What Is Marketing Analytics and Why You Should Be Gung-Ho About Its Future?

Previous ArticleNext Article

2 Comments

Leave a Reply

Your email address will not be published. Required fields are marked *